- Email: [email protected]
Metaanalysis of statins and survival in de novo cardiac transplantation
Metaanalysis of Statins and Survival in De Novo Cardiac Transplantation M.R. Mehra and N.Y. Raval ABSTRACT Background. The use of HMG CoA reductase inhibitors (statins) after cardiac transplantation has been suggested to decrease the incidence of severe rejection and improve survival. Individual investigations that have led to this suggestion are randomized (but not placebo-controlled) studies, including small patient numbers that have (and thus underpowered) and enrolling heterogeneous subjects (including retransplant recipients). The purpose of this pooled analysis was to quantify the benefit of statins on survival in de novo cardiac transplant recipients. Methods. Medline (1966 to 2003) was queried using the keywords statin, HMG CoA reductase inhibitors, cardiac transplantation, transplant, cholesterol, atorvastatin, fluvastatin, lovastatin, pravastatin, and simvastatin. In addition, we searched the cited literature and previously published systematic reviews. Of 36 articles retrieved, 3 randomized controlled studies met our population inclusion criteria; namely age ⬎18 years, de novo heart transplant recipients, statin therapy within 3 months, and ⱖ1-year follow-up. Pooled data were metaanalyzed by Mantel–Haenszel tests using a random effects model that included tests for heterogeneity. Results. The three pooled studies included 246 patients (statin, n ⫽ 129; no statin, n ⫽ 117) and 27 events (11%). The pooled analysis demonstrated a significant reduction in mortality with statin use (RR 0.31; 95% CI 0.13 to 0.7; P ⫽ .006) without significant heterogeneity (P ⫽ .7) among the studies. Two of the three studies reported allograft rejection with hemodynamic compromise. The pooled analysis demonstrated a significant benefit on this endpoint (RR 0.22, 95% CI 0.08 to 0.63; P ⫽ .004). Conclusion. This meta-analysis demonstrates that statin therapy decreases rejection episodes with hemodynamic consequences and improves 1-year heart transplant survival.
H
MG CoA reductase inhibitors (statins) have been suggested to improve survival and decrease both short- and long-term morbidity among cardiac transplants.1 Multiple clinical studies have examined mortality, rejection, and allograft vasculopathy among cardiac transplant patients utilizing statins. Individually, each trial has proposed that statin usage was beneficial for the endpoints of mortality and various morbid states, although uncertainties remain. First, although some studies are randomized, they are not placebo-controlled.2– 4 Second, these investigations have not consistently enrolled de novo heart transplant recipients and have included retransplant patients, who are known to have a natural history distinct from de novo cases.2 Other studies have focused primarily on composite endpoints because they lacked sufficient power to convinc-
ingly demonstrate a survival benefit, or they compared different statins without a control group.5,6 The purpose of this investigation was to perform a systematic review using metaanalysis to quantify the evidence linking statins to all causes of survival among recipients of first cardiac allografts.
From the Cardiomyopathy and Heart Transplantation Center, Ochsner Clinic Foundation, New Orleans, Louisiana, USA (M.R.M.); and Tulane University Health Sciences Center, New Orleans, Louisiana, USA (N.Y.R.). Address reprint requests to Dr Mandeep R. Mehra, Cardiomyopathy and Heart Transplantation Center, 1514 Jefferson Highway, New Orleans, LA 70121. E-mail: [email protected]
© 2004 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
0041-1345/04/$–see front matter doi:10.1016/j.transproceed.2004.05.036
Transplantation Proceedings, 36, 1539 –1541 (2004)
1539
1540
MEHRA AND RAVAL
METHODS Search Strategy The electronic database, Medline (Ovid), was searched from its year of inception to September 2003 using the keywords statin, HMG CoA reductase inhibitors, cardiac, transplantation, transplant, cholesterol, atorvastatin, fluvastatin, lovastatin, pravastatin, and simvastatin. Cited medical literature and previously published systematic reviews were also reviewed. The search strategy was limited to prospective, randomized study design, those published in the English language, and those investigating human subjects. Summary outcome data were extracted from the final trial reports.
Inclusion Criteria Our population inclusion criteria included: age ⬎18 years; de novo heart transplant; statin therapy within 3 months of transplant; at least 1-year follow-up in the statin-treated group; and presence of a control group. We initially selected reports independent of the inclusion criteria. Thirty-six potentially relevant articles were retrieved using a keyword search, bibliographic citation search, and review article searches. Of these 36 studies, only 3 randomized, prospective studies met our final inclusion criteria.2– 4
Data Abstraction All original articles were retrieved for data abstraction. Baseline demographic data, study design characteristics, and treatment effects were recorded by the two investigators, independently on a standard form. Because only de novo patients were assessed in the current analysis, those within the studies of interest listed as “second transplant” were removed from the evaluation. Where actuarial findings were presented in a multiple-year format, the Kaplan–Meier data were expanded to yield 1-year numeric data on the basis of survival curves. If an article reported different statins separately against a comparator statin-naive group, the statins were grouped into a single unit and results taken in aggregate.
Principle Endpoints The primary treatment endpoints included all causes of mortality at 1 year and, if reported, a secondary endpoint of acute rejection with hemodynamic compromise. Safety endpoints, including rhabdomyolysis and HMG CoA reductase inhibitor–associated hepatitis were also abstracted. Rhabdomyolysis was defined as CK-MB elevation greater-than ten times the upper limit of normal (ULN). Hepatitis was similarly defined as a liver function test (AST, ALT) abnormality greater than three times the ULN.
Statistical Analysis For purposes of the metaanalysis, the pooled data were analyzed using Mantel–Haenszel test in a random effects model that included tests for heterogeneity.
Fig 1. Pooled analysis of three randomized studies comparing heart transplant recipients administered statins vs those not administered statins. A significant survival benefit was detected in statin recipients.
was 179 minutes. Pooled analysis indicated no significant heterogeneity among patient characteristics, among the studies or between the statin and statin-naive groups including recipient age, donor age, recipient gender, ischemic etiology, ischemic time, total cholesterol, LDL cholesterol, and triglycerides. The medications were begun within 6 weeks after transplantation in the three studies. All three studies primarily treated patients with cyclosporine-based immunosuppression that included corticosteroids and either azathioprine or mycophenolate mofetil therapy. Primary Endpoint Analysis
A total of 27 deaths occurred within 1 year in the study cohort, representing an 11% 1-year mortality. The statintreated group had a 5.4% mortality (7 of 129), whereas the statin-naive group had a 17.1% mortality (20 of 117). Individual studies demonstrated consistent trends to improved survival, but in isolation neither study achieved statistical significance. Pooled analysis, however, demonstrated a statistically significant reduction in mortality associated with statin use (relative risk) [RR] 0.31; 95% confidence interval [CI] 0.13 to 0.7; P ⫽ .006), without significant heterogeneity (P ⫽ .7) among studies (Fig 1). No differences by type of statin were noted. Secondary Endpoint Analysis
Two of the three studies reported allograft rejection with hemodynamic compromise. Pooled analysis of this endpoint demonstrated a significant benefit of statin therapy on this endpoint (RR 0.22; 95% CI 0.08 to 0.63; P ⫽ .004). Safety Analysis
RESULTS Patients
The three studies that met inclusion criteria included 246 de novo heart transplant recipients. Of these, 129 received statins (pravastatin 20 to 40 mg/d, n ⫽ 68; simvastatin 10 to 20 mg/d, n ⫽ 61), whereas 17 patients were statin-naive. The mean age of the population was 50 years; 84% were men, mean donor age 31 was years, and mean ischemic time
No cases of rhabdomyolysis were reported in the three studies. Two of the three studies that evaluated liver enzymes reported no cases of transaminitis that met our criterion of a threefold elevation over baseline. One study reported two “excess” cases (12 cases in the statin group vs 10 in the statin-naive group) of elevated creatine kinase enzyme levels in the statin group (using simvastatin), but in all cases these were reversible.
STATINS AND SURVIVAL
DISCUSSION Study Findings
The results of our metaanalysis confirm the contribution of statins to the decrease in all causes of mortality at 1 year after de novo heart transplantation. Similarly, our study has also shown that statins may achieve some of this effect by influencing the endpoint of allograft rejection with hemodynamic compromise. However, the reliability of this conclusion is limited by the small number of events reported in only two of the three studies. More importantly, these beneficial effects are achieved with minimal adverse effects at the low doses of statins used in the trials. Mechanism of Benefits
Although initially conceived as lipid-lowering agents, statins have been subsequently shown to possess powerful immunomodulating properties.7 In the realm of transplantation, Katznelson and colleagues8 found that pravastatin suppressed natural killer cell cytotoxicity, a finding confirmed by Kobashigawa and colleagues,2 a subgroup of a clinical trial. Weis et al9 proposed that statin usage decreased the inflammatory milieu by modulating levels of cytokines (IL-6 and TNF-␣), which promote anatomic (decreased intimal thickness) and physiologic (coronary vasomotor function) endothelial health. Recent studies have alluded to an antiinflammatory effect of statins to reduce the levels of C-reactive protein, a biomarker associated with worse outcome in cardiac transplantation.10 In vitro experimentation has proven that statins interfere with cell growth signaling pathways through altered production of mevalonate and its downstream products—farnesylpyrophosphate and geranylgeranylpyrophosphate.3,11 Thus, it appears likely that the benefits of statins within the first year are due predominantly to their immunomodulatory and antiinflammatory effects, a proposition that is further endorsed by the suggestion that this therapy decreases hemodynamically compromising rejection episodes among cardiac allografts. Clinical Implications
Although individual studies have lacked sufficient power to quantify the benefits of statins, our analysis has demonstrated that statins are associated with 1 life saved for every 8.5 heart transplant recipients treated for 1 year. Looked at another way, statins might have the potential to save 471 lives each year among the 4000 de novo heart transplants performed worldwide. Even if we assumed a conservative estimate of benefit at the upper limit of the observed 95% CI, this would represent at least a 30% decrease in mortal-
1541
ity. Furthermore, the benefit of a pooled analysis is the possibility to detect an outcome benefit in disparate populations. Thus, the three studies analyzed included patients across broad geographic locations including ethnically diverse populations.2– 4 The findings provide reassurance for the widespread application of statins in de novo transplant recipients. One of the most important implications of this analysis is to interprete outcomes in clinical trials that evaluate newer immunosuppressive therapy. It is important to either stratify an outcomes analysis based on exposure to statins or to ensure that this therapy is considered mandatory. In our study we did not perform an analysis of the association between lipid-lowering effects and outcomes, because in most trials statin therapy was initiated regardless of baseline lipid levels, an approach that is recommended on the basis of our analysis. In conclusion, metaanalysis has demonstrated that statin therapy improves 1-year survival of de novo heart transplants and is associated with a decrease in severe allograft rejection episodes. REFERENCES 1. Ballantyne CM: Statins after cardiac transplantation: which statin, what dose, and how low should we go? J Heart Lung Transplant 19:515, 2000 2. Kobashigawa JA, Katznelson S, Laks H, et al: Effect of pravastatin on outcomes after cardiac transplantation. N Engl J Med 333:621, 1995 3. Mehra MR, Uber PA, Vivekananthan K, et al: Comparative beneficial effects of simvastatin and pravastatin on cardiac allograft rejection and survival. J Am Coll Cardiol 40:1609, 2002 4. Wenke K, Meiser B, Thiery J, et al: Simvastatin initiated early after heart transplantation: 8-year prospective experience. Circulation 107:93, 2003 5. Patel DN, Pagani FD, Koelling TM, et al: Safety and efficacy of atorvastatin in heart transplant recipients. J Heart Lung Transplant 21:204, 2002 6. Keogh A, Macdonald P, Kaan A, et al: Efficacy and safety of pravastatin vs simvastatin after cardiac transplantation. J Heart Lung Transplant 19:529, 2000 7. Mach F: Statins as immunomodulators. Transplant Immunol 9:197, 2002 8. Katznelson S, Wang XM, Chia D, et al: The inhibitory effects of pravastatin on natural killer cell activity in vivo and on cytotoxic T lymphocyte activity in vitro. J Heart Lung Transplant 17:335, 1998 9. Weis M, Pehlivanli S, Meiser BM, et al: Simvastatin treatment is associated with improvement in coronary endothelial function and decreased cytokine activation in patients after heart transplantation. J Am Coll Cardiol 38:814, 2001 10. Ventura HO, Mehra MR: C-reactive protein and cardiac allograft vasculopathy: is inflammation the critical link? J Am Coll Cardiol 42:483, 2003 11. Vincent TS, Wulfert E, Merler E: Inhibition of growth factor signaling pathways by lovastatin. Biochem Biophys Res Commun 180:1284, 1991
H
MG CoA reductase inhibitors (statins) have been suggested to improve survival and decrease both short- and long-term morbidity among cardiac transplants.1 Multiple clinical studies have examined mortality, rejection, and allograft vasculopathy among cardiac transplant patients utilizing statins. Individually, each trial has proposed that statin usage was beneficial for the endpoints of mortality and various morbid states, although uncertainties remain. First, although some studies are randomized, they are not placebo-controlled.2– 4 Second, these investigations have not consistently enrolled de novo heart transplant recipients and have included retransplant patients, who are known to have a natural history distinct from de novo cases.2 Other studies have focused primarily on composite endpoints because they lacked sufficient power to convinc-
ingly demonstrate a survival benefit, or they compared different statins without a control group.5,6 The purpose of this investigation was to perform a systematic review using metaanalysis to quantify the evidence linking statins to all causes of survival among recipients of first cardiac allografts.
From the Cardiomyopathy and Heart Transplantation Center, Ochsner Clinic Foundation, New Orleans, Louisiana, USA (M.R.M.); and Tulane University Health Sciences Center, New Orleans, Louisiana, USA (N.Y.R.). Address reprint requests to Dr Mandeep R. Mehra, Cardiomyopathy and Heart Transplantation Center, 1514 Jefferson Highway, New Orleans, LA 70121. E-mail: [email protected]
© 2004 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
0041-1345/04/$–see front matter doi:10.1016/j.transproceed.2004.05.036
Transplantation Proceedings, 36, 1539 –1541 (2004)
1539
1540
MEHRA AND RAVAL
METHODS Search Strategy The electronic database, Medline (Ovid), was searched from its year of inception to September 2003 using the keywords statin, HMG CoA reductase inhibitors, cardiac, transplantation, transplant, cholesterol, atorvastatin, fluvastatin, lovastatin, pravastatin, and simvastatin. Cited medical literature and previously published systematic reviews were also reviewed. The search strategy was limited to prospective, randomized study design, those published in the English language, and those investigating human subjects. Summary outcome data were extracted from the final trial reports.
Inclusion Criteria Our population inclusion criteria included: age ⬎18 years; de novo heart transplant; statin therapy within 3 months of transplant; at least 1-year follow-up in the statin-treated group; and presence of a control group. We initially selected reports independent of the inclusion criteria. Thirty-six potentially relevant articles were retrieved using a keyword search, bibliographic citation search, and review article searches. Of these 36 studies, only 3 randomized, prospective studies met our final inclusion criteria.2– 4
Data Abstraction All original articles were retrieved for data abstraction. Baseline demographic data, study design characteristics, and treatment effects were recorded by the two investigators, independently on a standard form. Because only de novo patients were assessed in the current analysis, those within the studies of interest listed as “second transplant” were removed from the evaluation. Where actuarial findings were presented in a multiple-year format, the Kaplan–Meier data were expanded to yield 1-year numeric data on the basis of survival curves. If an article reported different statins separately against a comparator statin-naive group, the statins were grouped into a single unit and results taken in aggregate.
Principle Endpoints The primary treatment endpoints included all causes of mortality at 1 year and, if reported, a secondary endpoint of acute rejection with hemodynamic compromise. Safety endpoints, including rhabdomyolysis and HMG CoA reductase inhibitor–associated hepatitis were also abstracted. Rhabdomyolysis was defined as CK-MB elevation greater-than ten times the upper limit of normal (ULN). Hepatitis was similarly defined as a liver function test (AST, ALT) abnormality greater than three times the ULN.
Statistical Analysis For purposes of the metaanalysis, the pooled data were analyzed using Mantel–Haenszel test in a random effects model that included tests for heterogeneity.
Fig 1. Pooled analysis of three randomized studies comparing heart transplant recipients administered statins vs those not administered statins. A significant survival benefit was detected in statin recipients.
was 179 minutes. Pooled analysis indicated no significant heterogeneity among patient characteristics, among the studies or between the statin and statin-naive groups including recipient age, donor age, recipient gender, ischemic etiology, ischemic time, total cholesterol, LDL cholesterol, and triglycerides. The medications were begun within 6 weeks after transplantation in the three studies. All three studies primarily treated patients with cyclosporine-based immunosuppression that included corticosteroids and either azathioprine or mycophenolate mofetil therapy. Primary Endpoint Analysis
A total of 27 deaths occurred within 1 year in the study cohort, representing an 11% 1-year mortality. The statintreated group had a 5.4% mortality (7 of 129), whereas the statin-naive group had a 17.1% mortality (20 of 117). Individual studies demonstrated consistent trends to improved survival, but in isolation neither study achieved statistical significance. Pooled analysis, however, demonstrated a statistically significant reduction in mortality associated with statin use (relative risk) [RR] 0.31; 95% confidence interval [CI] 0.13 to 0.7; P ⫽ .006), without significant heterogeneity (P ⫽ .7) among studies (Fig 1). No differences by type of statin were noted. Secondary Endpoint Analysis
Two of the three studies reported allograft rejection with hemodynamic compromise. Pooled analysis of this endpoint demonstrated a significant benefit of statin therapy on this endpoint (RR 0.22; 95% CI 0.08 to 0.63; P ⫽ .004). Safety Analysis
RESULTS Patients
The three studies that met inclusion criteria included 246 de novo heart transplant recipients. Of these, 129 received statins (pravastatin 20 to 40 mg/d, n ⫽ 68; simvastatin 10 to 20 mg/d, n ⫽ 61), whereas 17 patients were statin-naive. The mean age of the population was 50 years; 84% were men, mean donor age 31 was years, and mean ischemic time
No cases of rhabdomyolysis were reported in the three studies. Two of the three studies that evaluated liver enzymes reported no cases of transaminitis that met our criterion of a threefold elevation over baseline. One study reported two “excess” cases (12 cases in the statin group vs 10 in the statin-naive group) of elevated creatine kinase enzyme levels in the statin group (using simvastatin), but in all cases these were reversible.
STATINS AND SURVIVAL
DISCUSSION Study Findings
The results of our metaanalysis confirm the contribution of statins to the decrease in all causes of mortality at 1 year after de novo heart transplantation. Similarly, our study has also shown that statins may achieve some of this effect by influencing the endpoint of allograft rejection with hemodynamic compromise. However, the reliability of this conclusion is limited by the small number of events reported in only two of the three studies. More importantly, these beneficial effects are achieved with minimal adverse effects at the low doses of statins used in the trials. Mechanism of Benefits
Although initially conceived as lipid-lowering agents, statins have been subsequently shown to possess powerful immunomodulating properties.7 In the realm of transplantation, Katznelson and colleagues8 found that pravastatin suppressed natural killer cell cytotoxicity, a finding confirmed by Kobashigawa and colleagues,2 a subgroup of a clinical trial. Weis et al9 proposed that statin usage decreased the inflammatory milieu by modulating levels of cytokines (IL-6 and TNF-␣), which promote anatomic (decreased intimal thickness) and physiologic (coronary vasomotor function) endothelial health. Recent studies have alluded to an antiinflammatory effect of statins to reduce the levels of C-reactive protein, a biomarker associated with worse outcome in cardiac transplantation.10 In vitro experimentation has proven that statins interfere with cell growth signaling pathways through altered production of mevalonate and its downstream products—farnesylpyrophosphate and geranylgeranylpyrophosphate.3,11 Thus, it appears likely that the benefits of statins within the first year are due predominantly to their immunomodulatory and antiinflammatory effects, a proposition that is further endorsed by the suggestion that this therapy decreases hemodynamically compromising rejection episodes among cardiac allografts. Clinical Implications
Although individual studies have lacked sufficient power to quantify the benefits of statins, our analysis has demonstrated that statins are associated with 1 life saved for every 8.5 heart transplant recipients treated for 1 year. Looked at another way, statins might have the potential to save 471 lives each year among the 4000 de novo heart transplants performed worldwide. Even if we assumed a conservative estimate of benefit at the upper limit of the observed 95% CI, this would represent at least a 30% decrease in mortal-
1541
ity. Furthermore, the benefit of a pooled analysis is the possibility to detect an outcome benefit in disparate populations. Thus, the three studies analyzed included patients across broad geographic locations including ethnically diverse populations.2– 4 The findings provide reassurance for the widespread application of statins in de novo transplant recipients. One of the most important implications of this analysis is to interprete outcomes in clinical trials that evaluate newer immunosuppressive therapy. It is important to either stratify an outcomes analysis based on exposure to statins or to ensure that this therapy is considered mandatory. In our study we did not perform an analysis of the association between lipid-lowering effects and outcomes, because in most trials statin therapy was initiated regardless of baseline lipid levels, an approach that is recommended on the basis of our analysis. In conclusion, metaanalysis has demonstrated that statin therapy improves 1-year survival of de novo heart transplants and is associated with a decrease in severe allograft rejection episodes. REFERENCES 1. Ballantyne CM: Statins after cardiac transplantation: which statin, what dose, and how low should we go? J Heart Lung Transplant 19:515, 2000 2. Kobashigawa JA, Katznelson S, Laks H, et al: Effect of pravastatin on outcomes after cardiac transplantation. N Engl J Med 333:621, 1995 3. Mehra MR, Uber PA, Vivekananthan K, et al: Comparative beneficial effects of simvastatin and pravastatin on cardiac allograft rejection and survival. J Am Coll Cardiol 40:1609, 2002 4. Wenke K, Meiser B, Thiery J, et al: Simvastatin initiated early after heart transplantation: 8-year prospective experience. Circulation 107:93, 2003 5. Patel DN, Pagani FD, Koelling TM, et al: Safety and efficacy of atorvastatin in heart transplant recipients. J Heart Lung Transplant 21:204, 2002 6. Keogh A, Macdonald P, Kaan A, et al: Efficacy and safety of pravastatin vs simvastatin after cardiac transplantation. J Heart Lung Transplant 19:529, 2000 7. Mach F: Statins as immunomodulators. Transplant Immunol 9:197, 2002 8. Katznelson S, Wang XM, Chia D, et al: The inhibitory effects of pravastatin on natural killer cell activity in vivo and on cytotoxic T lymphocyte activity in vitro. J Heart Lung Transplant 17:335, 1998 9. Weis M, Pehlivanli S, Meiser BM, et al: Simvastatin treatment is associated with improvement in coronary endothelial function and decreased cytokine activation in patients after heart transplantation. J Am Coll Cardiol 38:814, 2001 10. Ventura HO, Mehra MR: C-reactive protein and cardiac allograft vasculopathy: is inflammation the critical link? J Am Coll Cardiol 42:483, 2003 11. Vincent TS, Wulfert E, Merler E: Inhibition of growth factor signaling pathways by lovastatin. Biochem Biophys Res Commun 180:1284, 1991